Toward a Sociocognitive Approach to Second Language Acquisition DWIGHT ATKINSON Graduate School of Education Temple University Japan 2-8-12 Minami Azabu Minato-ku Tokyo 106-0047 Japan Email: [email protected] This article develops the notion of a sociocognitive perspective on second language acquisition (SLA), proposed as an alternative to the cognitivism pervading the field. By sociocognitive, I mean a view of language and language acquisition as simultaneously occurring and interactively constructed both “in the head” and “in the world.” First, I develop a view of language and its acquisition as social phenomena—as existing and taking place for the performance of action in the (socially-mediated) world. Second, I describe the cognitive nature of language and its acquisition, focusing especially on recent developments in connectionism. Third, I introduce sociocognitive views of language and posit a social interpretation of connectionism as bridging the gap between cognition and social action. Fourth, I discuss sociocognitive perspectives on first language acquisition. Fifth, I describe the cognitivist biases of much SLA research, then suggest how sociocognitive approaches can help overcome them. I end by considering implications of the perspective I develop in this paper.

Theorists and researchers tend to view SLA as a mental process, that is, to believe that language acquisition resides mostly, if not solely, in the mind. (Davis, 1995, pp. 427–428) Most SLA researchers view the object of inquiry as in large part an internal, mental process. (Long, 1997, p. 319) SLA has been essentially a psycholinguistic enterprise, dominated by the computational metaphor of acquisition. (R. Ellis, 1997, p. 87) It is fair to say that the dominant theoretical influences [in SLA] have been linguistic and psycholinguistic. . . . While more socially oriented views have been proposed from time to time, they have remained relatively marginal to the field overall. (Mitchell & Myles, 1998, p. x) Much of what we identify as our cognitive capacities may . . . turn out to be properties of the wider, environmentally extended systems of which brains are just one (important) part. (A. Clark, 1997, p. 214) Individual effort and sociocultural activity are mutually embedded, as are the forest and the trees, and . . . it is The Modern Language Journal, 86, iv, (2002) 0026-7902/02/525–545 $1.50/0 ©2002 The Modern Language Journal

essential to understand how they constitute each other. Rather than according primacy to the role of sociocultural activity or of the individual, the aim is to recognize the essential and inseparable roles of societal heritage, social engagement, and individual efforts. (Rogoff, 1990, p. 25)

A RECURRING IMAGE COMES TO MIND WHEN I read much second language acquisition (SLA) research and theory. It is the image of a single cactus in the middle of a lonely desert—the only thing except sand for miles around. The cactus sits there, waiting patiently for that rare cloud to pass overhead and for that shower of rain to come pouring down. Like the solitary cactus, the learner in mainstream SLA research seems to sit in the middle of a lonely scene, and, like the cactus, the learner seems to wait there for life-giving sustenance (or at least its triggering mechanism)—input—to come pouring in. At that point the real action begins, and we watch the learner miraculously grow and change. A contrasting image sometimes also occurs to me, though more often when reading in fields other than SLA, such as language socialization and cultural anthropology. This is the image of a

526 tropical rainforest, so densely packed and thick with underbrush that it would be hard to move through. This forest is constantly wet with humidity and teeming with life, sounds, growth, and decay—a lush ecology in which every organism operates in complex relationship with every other organism. Each tree grows in and as a result of this fundamentally integrated world, developing continuously and being sustained through its involvement in the whole ecology. And this image satisfies me at a deeper level, because it corresponds to how I (and others) believe language acquisition “really works.” In this article, I undertake a critique of the “lonely cactus” view of SLA and offer in its place a perspective that integrates learners, teachers (not necessarily or perhaps even usually of the classroom variety), acquisitional contexts (both of situation and of culture, e.g., Halliday & Hasan, 1989; Malinowski, 1923; Ochs, 1990), and social practices, products, tools, and worlds (e.g., Berger & Luckmann, 1966; A. Clark, 1997; Gee, 1992; Wertsch, 1985). I argue that our obsession with the decontextualized, autonomous learner has prevented us from conceptualizing SLA as a situated, integrated, sociocognitive process—a viewpoint that will bear real fruit in attempts to understand the complex phenomenon of SLA. In order to reach this goal, however, one must start some way back. As SLA theorists (e.g., Gregg, 1988) have pointed out, studying SLA without first defining the substance and scope of the L and the A is a haphazard endeavor. Much of this article will therefore be devoted to establishing what I mean by the two foundational concepts language and its acquisition, as a necessary prerequisite to focusing the discussion explicitly on SLA. More specifically, the staging, sequencing, and relative emphasis of the exposition will be: (a) language and language acquisition as social phenomena, (b) language and language acquisition as cognitive phenomena, (c) language as a sociocognitive phenomenon, (d) language acquisition as a sociocognitive phenomenon; and (e) second language acquisition as a sociocognitive phenomenon.1 LANGUAGE AND LANGUAGE ACQUISITION AS SOCIAL PHENOMENA Obviously but nontrivially, language is social—a social practice, a social accomplishment, a social tool. People use language to act in and on their social worlds: to convey, construct, and perform, among other things, ideas, feelings, actions, identities, and simple (but crucial) passing

The Modern Language Journal 86 (2002) acknowledgments of the existence of other human beings. None of these activities makes sense apart from a fundamentally social environment—All language is language in use, to paraphrase M.A.K. Halliday. Language has brought manifold evolutionary advantages to our species, allowing us the possibility of both working smoothly and efficiently in groups, and critiquing and improving them. The social nature of language is not incidental to its existence and ongoing use—it largely if not completely explains them (e.g., Elman et al., 1996; Seidenberg, 1996). To the degree that we are cognitively predisposed to learn and use language, it is because as a social tool it allowed those who originally took advantage of it (in however rudimentary a form) an edge in survival over those who did not. And language as a species-wide capability continues to yield multiple advantages to human beings. In the defining moment of 20th-century linguistics, Chomsky took structuralism to its logical extreme, completely abstracting language from its social setting and declaring its ontological (or at least methodological) self-sufficiency. By reducing the social out of language he was able to produce an idealized pseudolanguage about which some “facts” could be explained using the tools of logic and calculus. Yes, speakers of English certainly do use question transformations—and arguably quite often do not, generally speaking, in oral discourse—but to base a whole linguistic theory on a handful of such phenomena belies the reductiveness and abstractness of Chomsky’s model of language.2 In fact, grammar (rightly understood) is itself a social accomplishment and social tool. Thus, a recent volume in Cambridge’s Studies in Interactional Linguistics series (Ochs, Schegloff, & Thompson, 1996) investigates the many ways in which grammatical features both function socially and are influenced by and shaped in interactional context: Grammar is part of a broader range of resources—organizations of practices, if you will—which underlie the organization of social life, and in particular the way in which language figures in everyday interaction and cognition. In this view, the involvement of grammar in such other organizations as those of culture, action and interaction has as a consequence that matters of great moment are missed if grammar’s order is explored as entirely contained within a single, selfenclosed organization. Grammar’s integrity and efficacy are bound up with its place in larger schemes of organization of human conduct, and with social interaction in particular.3 (p. 3)

Dwight Atkinson More specifically, the authors in this volume investigate how “grammar organizes social interaction” (p. 33; e.g., how grammatical structure contributes to conversational turn-taking); how “social interaction organizes grammar” (p. 36; e.g., how grammatical structure varies, changes, and is emergent across social settings and sociohistorical time); and how “grammar is a mode of social interaction” (p. 38; e.g., how people coconstruct utterances using grammar as a shared resource). Beyond grammar, a brief sampling of phenomena which socially-oriented linguists have attempted to account for—and which, I would argue, deserve a central place in any valid, comprehensive approach to language and its acquisition—include: 1. Politeness, identity, and presentation of self. A major use of language is to negotiate and maintain relationships between people. This includes its central role in presenting and performing identities, or socially expressive versions of the self (e.g., Brown & Levinson, 1987; Gee, 1990; Goffman, 1959; Peirce, 1995; Tannen, 1986). 2. Perspective taking and contextualization cueing. All language in use incorporates markers of how it is to be interpreted (e.g., Gumperz, 1982; Quirk, Greenbaum, Leech, & Svartvik, 1985). Thus, the string I hate you, while not devoid of (socially-structured) lexical and grammatical meaning, would usually be incomprehensible as a spoken utterance without at least some of the following (which typically co-occur): intonation, loudness, voice quality, and emphasis (but see next point).4 3. Language-in-context. Of central importance to a truly social understanding of language is what else goes on vis-à-vis the languaging event that contributes to its situated significance. To complicate usefully the example given to support the previous point, if the I hate you is accompanied by a passionate embrace, changes in the intonation contour may not be needed to gauge the intent of the statement. Language never occurs apart from a rich set of situational/sociocultural/historical/existential correlates, and to separate it out artificially is to denature it. Recent research on the complex implicatedness of speech with gestural systems, eye gaze and head movement, bodily orientation, and the manipulation of objects (e.g., Goodwin, 2000; Goodwin & Goodwin, 1992; Kendon, 1992; McNeill, 2000; Ochs et al., 1996) is highly suggestive in this regard. Discourse analysts and anthropologists (e.g., Fairclough, 1992; Gee, 1990; Hutchins, 1995;

527 Pennycook, 1994, 1998; Street, 1984, 1993) have also dealt with the profound sociohistorical imbrication of written and spoken language in systems of social and cultural practice, hierarchy, professional specialization, and power. 4. Turn-taking, participation structures, and opportunity structures. The mechanisms and ideologies by which participation in language activity is socially apportioned are central to the notion of language as a social phenomenon (e.g., Sacks, Schegloff, & Jefferson, 1974). They also vary markedly according to cultural norms of language use and the distribution of power in society (e.g., Bourdieu, 1991; Edelsky 1981; McDermott, 1988; Ochs, 1988a; Philips, 1983; Tannen, 1993; Tannen & Saville-Troike, 1985). 5. Speech as an interactional accomplishment. A number of scholars (e.g., Goodwin, 1986, 1987, 2000; Lerner, 1996; Ochs, Schieffelin, & Platt, 1979) have shown how oral language is structured across individuals, rather than by individuals operating autonomously. A persistent contention has been that the individual is not the appropriate unit of analysis when examining language and associated behavior from a truly social point of view (e.g., Bakhtin, 1990; Rogoff, 1990, 1998). 6. Social indexicality. Overlapping and possibly subsuming several of the categories already mentioned, the notion of indexicality, broadly construed, suggests the all-important use of language to orient oneself and others in the (socially-mediated) world. Far beyond the perceived oddities of traditional indexicals—e.g., personal pronouns and deictic expressions—social indexicality sees virtually all linguistic referrings as underspecified, and, therefore, as taking their meaning as much from their contextual (and sociocontextually-construed) surroundings as from their literal sense (Hanks, 1996; Ochs, 1990, 1992, 1996). Other phenomena that a valid approach to language-in-the-world would have to account for, but which cannot be explicated here, include: social knowledge of and participation in speech events, sociolinguistic (including register) variation, and the organization and “addressivity” (Bakhtin, 1990) of discourse. None of the linguistic features or functions mentioned to this point should be considered add-ons to a basic framework for understanding language—on the contrary, they comprise part of the core set of phenomena that such a framework needs to explain. Just as surely as language is social, so is its acquisition. As with other social practices, new-

528 borns are actively inducted into “languaging” (Becker, 1988) from day one, or almost certainly before (Foster, 1990; Locke, 1993, 1995). This is not to suggest that infants are not cognitively predisposed to language-related phenomena— to human voices and faces (e.g., Bower & Wishart, 1979; Eimas, 1975), for instance, or pronounced intonation contours (Snow, 1995)—or that children themselves are anything less than active participants in the first language (L1) acquisition process. What it does suggest is that such cognitive potentials are realized in (and such dynamic actors supported by) extremely rich, nurturing social activities and contexts. Quite opposed to the Chomskyan “poverty of the stimulus” argument, a first principle of L1 acquisition should therefore be the richness of the context, that is, the deep, multiplex embedding of language activities in the lush social world that surrounds most children, and which, in the words of Bourdieu (1991), “instead of telling the child what he [sic] must do, tells him what he is, and thus leads him to become durably what he has to be” (p. 52). Thus, whereas communicative intentionality as we know it cannot be attributed to newborns and early infants, caregivers often interpret their behaviors as intentional (Foster, 1990; Smith, 1988) and act on that basis. As Newson (1979, cited in Foster 1990) expressed it: “Human babies become human beings because they are treated as if they already were human beings [italics added]” (p. 16). A second principle of L1 acquisition should therefore be the social input (or insertion—Gee, 1995) principle. In opposition to the notion of input as linguistic information presented to and activating the language learning automaton (i.e., the lonely cactus view), input is used here to denote that the child herself is input/inserted into an ongoing stream of social interaction that supports her language development at every turn. Gee (1995) generalized his nearly identical “insertion principle” of L1 acquisition to all kinds of complex learning: The insertion principle: Efficacious learning of a new complex system is a process involving socially supported and scaffolded insertion into an activity that one does not yet understand. (p. 336)

Gee (1995) also described additional principles by which L1 acquisition takes place vis-à-vis its rich social environment. Among them are: The routine principle: Early insertion into an activity one does not yet understand requires that the activity be to a certain extent repeated and routinized or

The Modern Language Journal 86 (2002) ritualized. Rituals “freeze” meaning for the learner’s observation. (p. 336) The public principle: The meanings of the parts of new systems, whether words, visual symbols, actions, or objects must initially be rendered public and overt, so that the learner can see the connection between the signs and their interpretations. And this is done, in first language acquisition and other forms of learning, by the ways in which words, actions, and social interaction are integrally intertwined. (p. 337) The context-variability principle: In learning the parts of a new system, the learner will initially tie meanings to specific contexts or experiences. Appreciating wider meaning is a matter of having multiple experiences, not (just) learning “general rules,” and mastery requires practice at varying aspects of meaning so as to actively fit them to the context of use. People who know only “general meanings” and cannot vary these in context neither know the system, nor are they acquiring it in a useful way. (p. 346)

A substantial literature on child language socialization now exists that details the manifold, diverse ways in which language is shaped for and by acquirers into the dynamic, creative social signifying system that it is. The basic assumption underlying socialization in general is that “children come to share the world view [and social practices] of their community through the arrangements and interactions in which they are involved, whether or not such arrangements and interactions are intended to instruct them” (Rogoff, 1990, p. 98). Consequently, early language socialization studies focus on language learned “through intensive . . . contact under conditions allowing maximum feedback such as we find in home and peer settings” (Gumperz, 1982, p. 139). Given the highly active stimulus-seeking nature of children, this makes the “conservative[ly]” estimated 12,000–15,000 hours of intensive contact between average caregivers and children over their first years together (Larsen-Freeman, 1991, p. 336; for similar estimates see N. C. Ellis, 1998; McLaughlin, 1987) a period in which a seemingly impossible amount of learning-in-context can take place. Although language is clearly internalized in a sense during L1 acquisition, however, it never ceases to be part of the learner-as-social member’s set of interactively constructed social tools, practices, and experiences, and, in this way and others, continues to be held jointly with the social world. In fact, I will argue below that too much has been made of the internal/external opposition—the division between the cognizing individual and the social (or socially-mediated) world.

Dwight Atkinson LANGUAGE AND LANGUAGE ACQUISITION AS COGNITIVE PHENOMENA To say that language is social is in no sense to deny that it is also cognitive. Humans have evolutionarily highly-developed brains, and language, to some substantial degree, is stored in, comprehended by, produced by, and therefore reflects the basic design features of the human brain. In fact, the human cognitive apparatus is much more of a virtuoso language performer than rather cutand-dried functional portrayals typically allow (although, as will be explained further below, its effective performance can only take place by virtue of a rich, interactive fund of social knowledge, in a rich social/interactional context). Thus, in the act of conversing, we not only store and retrieve linguistic and contextual information “online” almost instantaneously and normally with great accuracy, but we also produce, comprehend, monitor our production and comprehension, plan our next contribution, and perform myriad other operations (e.g., adopt various politeness strategies, negotiate turn-taking, anticipate what others are going to say, integrate different modes of language and symbol system use such as reading, gesturing, eye gaze, and body orientation with speaking) virtually all at the same time. It is small wonder then that some psychologists have begun to redefine intelligence in terms of the amazing things all (or at least many) humans do cognitively—including, notably, in their use of language—rather than in terms of intelligence measures which place individuals in rigid hierarchies of difference and deficit (Scribner & Cole, 1981; cf. Gee, 1990; Gould, 1981). Several recent cognitive theories of language and grammar have presented interesting contrasts to Chomskyan linguistics in the important roles they give to context and use. Thus, Hopper (1988) coined the term emergent grammar to describe a view of grammar as “a vaguely defined set of sedimented (i.e., [more or less] grammaticized) recurrent partials whose status is constantly being renegotiated and which cannot be distinguished in principle from stategies for building discourses” (p. 118). He went on to give evidence from written Malay of the ways such discourse phenomena as foregrounding and backgrounding substantially shape the grammar that is used to enact them. Other scholars proposing cognitively focused but context-driven and noncomponential approaches to grammatical knowledge include Becker (1979; 1988), Langacker (1987), C. J. Fillmore, Kay, and O’Connor (1988), and Seidenberg (1996). Their theories all have in common the notion that gram-

529 mar does not represent a single, logical, a priori system whose basic existence and implementation for communicative purposes is irrelevant to its cognitive status. Language acquisition as well is obviously a cognitive phenomenon. Cognition and language develop hand in hand in early childhood, and less dramatically into the later years—in several important senses their development never ceases (e.g., Kemper, 1987). Cognitive predispositions toward learning and using language (Foster, 1990; Locke, 1993, 1995) are clearly also present at the beginning of life—infants truly do seem to come equipped with either linguistic protoknowledge already built in, or cognitive systems primed to learn an amazing amount linguistically in a short time. Operating principles, parameter setting, bootstrapping, and child-directed speech are all attempts to account for the incredible amount of language awareness children display early in life, and for how they so quickly complexify it. One recent cognitive attempt to explain L1 acquisition is connectionism (e.g., Plunkett, 1995; Rumelhart & McClelland, 1986; cf. N. C. Ellis, 1998). Connectionist theories depend on the computational modeling of language (and other kinds of) learning via the gradual buildup of richly internetworked association potentials at the neural level which are selectively and simultaneously activated in specific patterns to perform cognitive activities, including language production and comprehension.5 Connectionist L1 acquisition models appear to account for how various complex language systems such as the English past tense verb-marking system (e.g., Plunkett & Marchman, 1993; Rumelhart & McClelland, 1986) can be learned over time without assuming innate linguistic knowledge. Although much effort has gone into establishing the modeling of this particular grammatical system as a paradigm example of connectionism’s power, connectionist researchers have investigated other linguistic domains as well. Thus, Elman (1992) and others have modeled connectionist systems that correctly assign syntactic category labels after exposure to a range of syntactic strings, and Stemberger (1992) “has offered a connectionist model as a plausible account of the characteristics of child phonology” (Leonard, 1995, p. 590). A major strength of connectionist approaches to L1 acquisition is that, unlike their innatist counterparts, they appear to account for empirically determined hallmarks of the acquisitional process. These hallmarks include: in the acquisition of the lexicon, semantic overextension and

530 underextension (e.g., Barrett, 1995; Gruendel, 1977) and the “vocabulary burst” often seen in children during their second year (e.g., Bates, Dale, & Thal, 1995; Nelson, 1973); overgeneralization of past tense verb endings (e.g., Brown, 1973; Marchman & Bates, 1994); the comprehension-production lag in language acquisition (e.g., E. V. Clark, 1983; Plunkett, 1995); and U-shaped growth in a number of areas (e.g., Brown, 1973; Plunkett, 1995). Just as importantly, perhaps, connectionist researchers have modeled interdomain language acquisition. Thus, Bates et al. (1995) discussed the efficacy of connectionism in accounting for the “continuity hypothesis” (cf. Foster, 1990) relating early lexical development to early grammatical development in English. They pointed out that in the past tense morphology-learning simulations mentioned above there is “temporal asynchrony” between the acquisition of discrete lexical forms and the emergence of a rule-like system that correctly assigns past tense inflectional endings to previously unseen verbs: In the early stages of learning, the system appears to learn each mapping from present to past tense by rote, with no generalization to novel lexical forms. . . . As instances of present/past tense mapping accumulate, some dramatic nonlinear changes are observed: the rate of learning accelerates markedly, overgeneralization errors start to appear, and the system starts to provide a default mapping to novel items. . . . Superficially, the network behaves as though it has switched from one mode of learning (rote) to another (rule). And yet there are no structural discontinuities in the system itself, or in the one-verb-at-atime nature of the input. . . . Simply put, grammatical generalizations (i.e., rulelike behaviors) do not arise until this system has acquired enough instances to support those generalizations. When the requisite number of items has been acquired, dramatic changes can take place, even within a single system. (Bates et al., 1995, pp. 118–119)

Connectionist approaches to language acquisition thus have the potential to tell us much about the cognitive mechanisms implicated in L1 acquisition. In addition, they have characteristics—to be discussed in the following two sections of this article—that allow the cognitive and social aspects of language use and acquisition to be tied much more closely together than previously. LANGUAGE AS A SOCIOCOGNITIVE PHENOMENON In describing the social and cognitive attributes of language to this point, I have purposely tried

The Modern Language Journal 86 (2002) to keep them separate. In actual fact, however, a growing number of linguists, anthropologists, developmental psychologists, and language acquisition researchers do not accept such a dichotomy (e.g., Gee, 1990, 1992; Halliday, 1978; Hanks, 1996; Lave & Wenger, 1991; Ochs, 1992, 1996; Rogoff, 1990).6 Along with these scholars, I consider the social and cognitive aspects of language to have co-evolved from the beginning, and therefore to function interdependently—if not inseparably. In this section, I will try to describe how the social and the cognitive work together in this way. A number of linguistic phenomena can, in fact, only be accounted for if the cognizing individual’s linguistic knowledge is seen to be abetted by, actuated within, and broadly continuous with a rich social context. Researchers (e.g., Goodwin, 1986, 1987; Lerner, 1993, 1996; Ochs et al., 1979), for example, have described the interactional accomplishment of propositions across individuals—the remarkable ability of individuals to contribute jointly to the expression of a single idea or action. Such a feat is possible only in the presence of joint cognition—a form of intersubjectivity created and maintained on the basis of both shared (and highly-articulated) cognitive knowledge of the activity being engaged in, and a world that gives such activity a social purpose, a conventional shape (e.g., a participation structure), and an approximately agreed-upon means of linguistic expression. Among socially-oriented linguists, conversation is frequently considered the paradigm speech activity within the human language-making capacity (e.g., Ochs et al., 1996). One of the foremost characteristics of conversation is its jointly-accomplished nature, whether within or across utterances. Thus, topic nomination by one party often leads to multiple utterances on the topic (which itself may be subtly negotiated and modified in the interaction) by both the nominator and other parties, leading eventually to the expression of a broadly-shared perspective, if only for the duration of the interaction (e.g., Goodwin, 1981, 1986, 1987). Likewise, conversational storytelling is a highly co-constructed sociocognitive activity, with different parties exercising influence over the story’s course through their responses, by sharing more actively in its co-construction, or even by taking it over as it proceeds (Goodwin, 1986, 1987; Mandelbaum, 1987).7 The effective use of any language, in fact, crucially assumes the preexistence and interactionally-achieved development of shared sociocognitive perspectives—thus a basic precondition for effective lin-

Dwight Atkinson guistic reference (e.g., nominal, deictic, or definite reference—H. H. Clark & Marshall, 1981; H. H. Clark & Wilkes-Gibbs, 1986) is that language users actively co-construct and adopt a common referential ground (Hanks, 1996).8 It is necessary at this point to consider in more detail how the cognizing individual and the socially-constructed and mediated world make their integrated contributions to linguistic activity in that world. In order to do so, I will adopt a connectionist understanding of cognition, but one that extends rather directly into the social world, and vice versa. My account substantially follows that of Gee (1992). As described previously, connectionism posits that meaning/knowledge exists largely in potential form in the human cognitive apparatus. That is, rather than having prebuilt cognitive structures, or schemas stored in the brain, all that exists is the potential for such structures to be formed basically online through the activation of various networks of neural associations. Thus, meaning (or its material substrate) is distributed across a large number of neurons, all of which have varying probabilities of firing in concert with other neurons, depending both on the nature of the input stimulus and on what has been experienced in the past—and thus what connections have been previously trained/socialized to activate together. For instance, while walking along a New England country road on a sunny summer day, a passing stranger’s “Hi!” or “Nice weather we’re having”—suitably accompanied by eye contact, a smile, a certain affective disposition (at least to the extent of desiring to communicate), the orientation of the speaker’s face toward that of the (assumed) recipient of the utterance, delivered in a suitably friendly tone at a socially acceptable distance while enacting a socially appropriate identity, and so forth—will, in principle, activate in the recipient a set of previously socialized neural connections leading to the holistic understanding that this is the speech act we call a greeting. Quite probably, it will also result in neurally-based and highly-patterned action by the recipient, comprising, among other things, a reciprocal physical orientation, affective stance, facial expression, tone of voice, and utterance—what we call a response. What I first want to highlight in this example is the interactive, “outside-in” (Shore, 1996) nature of the sociocognitive event. Without the rich contribution of social (signifying) practices and tools such as facial expression, physical orientation, voice quality, language, social scene (an isolated New England byway), and social actors (strangers

531 approaching each other), neither cognition itself nor the resulting speech event could have taken place. Put a bit more metaphorically, perhaps, the acts of cognition described in this example are substantially continuous with the social world—they do not start in the head, although the head is certainly involved, nor do they end in the head, because the output is social action. Nor do the social (signifying) practices involved simply take on their meaning once they arrive in the head; instead, they come with meaning already, in a sense, built in—just as language carries with it meaning that is only “borrowed” (Bakhtin, 1990) in specific instances of language use. The point here, then, is that cognition is not a private activity that occurs exclusively in the confines of an independent, isolated cerebral space, but rather that it is at least a semipublic activity, produced as part of a substantially open system. Whenever we participate in social activity, we participate in conventional ways of acting and being that are already deeply saturated with significance. Gee (1992, p. 12) gave the following example of how cognition (or at least knowledge) is both “in the head” and “in the world”: Consider the way people move around a city. Some people undoubtedly have quite impressive “maps” of the city in their heads, others have less complete ones, and some people have quite impoverished ones. However, people do not need to have any very full representations in their minds . . . since the structure of the city, out in the world as it is, determines a good deal of their movement. . . . People’s “knowledge” of the city is stored, in good part, out in the city itself. Their “city schema” . . . is not just made up of things in their heads, it is also composed of the structures in the city itself, as well as physical maps (and things like public transportation schedules) that people can read. (p. 12)

Thus, meaning resides (partway) in social products (e.g., cities, maps, country roads, cars), social practices (e.g., greeting somebody, reading, baking a cake, playing the role of teacher or student in a classroom), and social tools (e.g., language, literacy, computer programs, methods of navigation), even as it resides (partway) in the head. A second, slightly different aspect of the country road greeting example that I want to highlight is the profoundly integrative nature of the sociocognitive event. That is, language activity and its cognitive correlates always occur as integral parts of larger sociocognitive wholes. Thus, without each of the following occurring and being cognized in relation to one another—socially signify-

532 ing facial expressions and physical orientations, affective stance, a conventionalized social scene in a social setting with a social purpose performed by social actors, and the effective deployment of the social tool of language—the activity of exchanging greetings could not have taken place. An important strength of connectionism is that it can account for such complex intertwinings of sensory and affective modalities, emergent cognitive structures, social phenomena, and language ability and use in the accomplishment of social activity. It does so in terms of the interactions of various sets of activated neurons—of neural networks that work together because of previous exposure to similar experience complexes (e.g., Gee, 1992, chap. 2; Strauss & Quinn, 1997, chap. 3). From this viewpoint, language per se is not the privileged and separate summum bonum, but rather an integral element of sociocognitive activity, the ultimate purpose of which is to perform situated action-in-the-world. LANGUAGE ACQUISITION AS A SOCIOCOGNITIVE PHENOMENON Equally good arguments for the sociocognitive nature of language come from the circumstances of its acquisition. A well-established finding in L1 acquisition research is that external context plays a crucial role early on. In early acquisition of the lexicon, for instance, it is vital that children be able to link particular sounds, words, and phrases repetitively to the appearance of particular objects or to the occurrence of particular actions (Barrett, 1995; see also Gee, 1995, described previously). To say that linguistic form is simply being “input” into an autonomous language learning system at this point would be inaccurate; at minimum, both the form and the object/action are being cognitively represented in some kind of associative relationship. The widely accepted notion of event representation (e.g., Barrett, 1995; Nelson, 1986) goes one step further, providing a richer and, from a sociocognitive perspective, more plausible view of early language acquisition. It postulates that children already have cognitively well-developed, organized associations of activity sequences, actors, and objects (along with open slots by which some actors and objects may be substituted for by others) onto which different words are initially mapped, and within which they thereafter become substantially integrated. Nor is it just any external context that early language development depends on. Rather, socialinteractional routines (e.g., Barrett, 1995; Gee,

The Modern Language Journal 86 (2002) 1995) provide both the framework and the content through which basic context-language form associations are made and internalized. Gee’s (1995) routine principle and public principle, described earlier, capture important facets of how such routines provide repetitive, public, and highly-structured opportunities for infants to connect specific utterances with specific semantic complexes. Coordinated activity of infant and caregiver is central in this kind of routine—despite the fact that intentionality cannot be strictly attributed to infants, as aggressively stimulus-seeking organisms with a natural attraction to interactive and repetitive behavior (Locke, 1993) they learn to play their part early on. More specifically, the related notions of scaffolding (e.g., Ninio & Bruner, 1978), vertical constructions (Scollon, 1976), zone of proximal development (e.g., Cole, 1985; Vygostky, 1978), socializing attention (Zukow-Goldring & Ferko, 1994), and propositions across individuals and utterances (Ochs et al., 1979) all reconceptualize the child and the caregiver as an interactional unit in their accomplishment of sociocognitive tasks that the child could not perform independently, and, therefore, as a central mechanism of developmental growth. It is generally thought that as L1 development progresses, context becomes a less important part of the acquisitional scheme. Thus, whereas children’s early utterances are tied closely to the here and now (Foster, 1990), they become progressively less so as time goes on. This finding is often taken as evidence that some sort of basic language learning machinery has kicked fully into gear and is now operating on a fairly systematic and autonomous basis. Although it is no doubt true that qualitative changes take place once linguistic knowledge reaches a certain critical mass for children, connectionist theories of cognition and language acquisition suggest that much if not all learning is accomplished not by progressively greater separation of the knowledge being acquired from the external world and other domains of cognitive knowledge, but by their increasing integration. If this is the case, then it might be more accurate to say that context, rather than disappearing from the scene altogether, partly “comes inside.” That is, as has been widely posited by neo-Vygotskian sociocultural theorists (e.g., Lantolf & Appel, 1994; Wertsch & Stone, 1985), the fundamental dynamic of acquiring language (as well as other “higher mental functions”) is that formerly externalized/social knowledge is substantially reconfigured as internalized/cognitive knowledge. But scholars who have recently attempted to link cog-

Dwight Atkinson nition via connectionism more closely to the social and cultural world (e.g., Gee, 1992; Shore, 1996; Strauss & Quinn, 1997) might take exception to the implications of the Vygotskian claim that language is, to some degree, gradually loosened during development from its moorings in social life. Rather, from this newer perspective, it might be more accurate to say that social life is now becoming articulated at a cognitive level in the development of cultural models (Gee, 1992; Holland & Quinn, 1987; Shore, 1996)—models that owe both their origins and their continued activation and use directly to their social existence—and that in this important sense the cognitive and the social are growing progressively closer together. If this version of connectionism is even approximately correct, then the notion of “decontextualized language”—so important in descriptions of later language and literacy learning and early success and failure in schooling—is perhaps neither useful nor accurate. That is, if context is taken to be only the immediate social and physical setting of language behavior in the here and now—the context of situation first described by Malinowski (1923) and later elaborated by Halliday (in Halliday & Hasan, 1989, chap. 1)—such language behavior might well be seen as thoroughly decontextualized. But if context is more widely construed as context of culture (Halliday & Hasan, 1989; Malinowski, 1923; Ochs, 1990)—involving the increasing cognitive interarticulation of cultural models with social practices, social products, and social tools (including, crucially, how to deploy them effectively)—then to think of language as becoming progressively more decontextualized and autonomous in the course of its acquisition may be fundamentally misguided. Event representations in early L1 acquisition, although sometimes claimed to be relatively temporary and evanescent, suggestively resemble early and relatively unarticulated schemas or mental models (or, in connectionist terms, networks of neural associations out of which such schemas/models are formed basically online). Whether or not they provide a skeletal base for the later articulation of sociocultural knowledge in the form of cultural models, schemas, frames, or scripts (Cole, 1985), their early existence does suggest a sort of continuity of sociocognitive processes—the progressive “thickening” of “knowledge in the head” by “knowledge in the world”—the basic outside-in dynamic described by Shore (1996) and alluded to above. Put yet another way, any organism’s progressive adaptation to its environment is a hallmark of develop-

533 mental growth. The gradual approximation by developing humans of what is “in the head” to what is “in the world” is just such an adaptive dynamic. Gee (1992; following C. J. Fillmore, 1975) described the cultural model associated with the lexical item bachelor. According to Gee, the dictionary definition of bachelor as unmarried male is only rudimentarily related to the dense linkings of associations that fall within the culturally construed understanding of bachelor. For one, bachelor invokes an idealized, culturally normative view of masculine sexuality. Thus, Roman Catholic priests, young boys, homeless men, and so forth, seem to fit into the category only problematically. Likewise, when contrasted with the term spinster—its apparent gender counterpart—bachelor seems to carry with it the cultural presupposition that single men become progressively more eligible and desirable marriage partners as they grow older (or at minimum their eligibility never ceases), whereas women become progressively less so. Although not all connections in the complex web of potential associations with the word bachelor need be activated on every occasion of its use, the point is that such a rich, articulated background exists, and that this background is located not only in cognitive but also social space. It might also be pointed out that examples of the type I have given in this article cannot even begin to approximate the extremely rich, complex, and highly-articulated kinds of sociocognitive knowledge that most humans are privy to simply by virtue of their lived experience and sociogenetic inheritance. To mention just one further (and more or less linguistic) element that must figure importantly in such highly developed knowledge complexes vis-à-vis language use, consider the growing realization among linguists that statistical-collocational relationship between various linguistic features is a major principle in the syntagmatic and paradigmatic organization of language-in-use and its acquisition (e.g., Saffran, Aslin, & Newport, 1996; Seidenberg, 1996; Stubbs, 1996, chap. 2). Such knowledge is easily accounted for in terms of connectionist networks (Seidenberg, 1996), and would obviously be held in joint, if variable, ownership with one’s social milieu. Taken together with the other kinds of knowledge I have so far mentioned and exemplified, it suggests “a semantics . . . more like an encyclopedia than a dictionary, [which] incorporates the view that linguistic categories exist in relation to ‘particular structured understandings of cultural institutions, beliefs about the world’” (Hanks, 1996, p. 244, including a quotation from

534 C. J. Fillmore, 1985). Although this formulation may still assume something of the traditional divide between the social and the cognitive I have been arguing against, it also captures some of the interpenetration and connectivity I am trying to portray.9 SLA AS A SOCIOCOGNITIVE PHENOMENON Let me summarize the main argument of the article as developed to this point: Neither language acquisition nor language use—nor even cognized linguistic knowledge—can be properly understood without taking into account their fundamental integration into a socially-mediated world. Beyond simply saying that the cognitive and the social interact, I use recent research in linguistics, anthropology, language acquisition, and cognitive science—work I have necessarily had to rely on, because so little of its kind has yet been done in the field of SLA—to argue that they are mutually constituted. It is commonplace for cognitively-oriented and socially-oriented language researchers to assume (at least for working purposes) the separability of their domains. I argue instead that these two domains cannot usefully be separated: The coinage “sociocognitive” (e.g., Ochs, 1988b) is adopted to represent this distinct perspective. With its beginnings in the cognitive psychology revolution of the 1960s (e.g., Brown, 1973; Corder, 1967), the field of SLA has adopted, by and large, a highly cognitivist view of second language learning (e.g., Breen, 1985; Crookes, 1997; Firth & Wagner, 1997). By cognitivist, I mean a perspective that places SLA mainly within individual heads and that sees individuals therefore as radically autonomous language acquirers (Pennycook, 1997). Even where social variables appear to enter in, from this perspective, they do so only as indirect influences on the acquisitional process (R. Ellis, 1994).10 Although this shared perspective has provided a (very) roughly common ground for inquiry in SLA, it comes at the cost of denaturing reduction. If the development of “languaging” depends on greater engagement with and adaptation to the (socially-mediated) world—or, more accurately, on the progressive interarticulation of the social and the cognitive—then a SLA based substantially on such master concepts as input, the (idealized) learner, and a “lonely” version of cognition is an impoverished endeavor. Others have, of course, made similar points from time to time, and in recent years mainstream SLA has been increasingly criticized for

The Modern Language Journal 86 (2002) this reason (as seen, in part, in the quotations beginning this article). But Breen’s (1985) plea to researchers and teachers to treat second language classrooms as cultural scenes is still one of the most eloquent statements of the problem: Given that we wish to understand how the external social situation of a classroom relates to the internal psychological states of the learner, the metaphor of the classroom as provider of optimal input or reinforcer of good strategies is inadequate. It reduces the act or experience of learning a language to linguistic or behavioural conditioning somehow independent of the learner’s social reality. Not only is SLA research currently offering us a delimited [sic] account of language learning, reducing active cognition to passive internalisation and reducing language to very specific grammatical performance, the mainstream of SLA research is also asocial. It neglects the social significance of even those variables which the investigators regard as central. The priority given to linguistic and mentalistic variables in terms of the efficient processing of knowledge as input leads inevitably to a partial account of the language learning process. The social context of learning and the social forces within it will always shape what is made available to be learned and the interaction of individual mind with external linguistic or communicative knowledge. Even Wundt, the first experimental psychologist, believed that he could not study higher mental processes such as reasoning, belief, thought, and language in a laboratory precisely because such processes were rooted within authentic social activity. (pp. 138–139)

Likewise, Vygotskian researchers (e.g., Hall, 1997; Lantolf, 2000; Lantolf & Appel, 1994) have argued against the reductive cognitivism of the field. Lantolf and Appel (1994), for example, criticized a view of the language learner “as a solipsistic biological organism whose cognitive powers simply unfold or ripen with the passage of time, rather than as someone who experiences productive participation in joint activity” (p. 11). Discourse analysts Firth and Wagner (1997) also offered a broad critique of SLA research as “individualistic and mechanistic”—as based on a perspective that is: weighted against the social and the contextual, and heavily in favour of the individual’s cognition, particularly the development of grammatical competence. This has led to an imbalance of adopted theoretical interests, priorities, foci, methodologies, perspectives . . . resulting in distorted descriptions of and views on discourse, communication, and interpersonal meaning—the quintessential elements of language. Moreover, this has occurred even in SLA work that is concerned with discourse and interaction. (p. 288)

535

Dwight Atkinson Although I do not agree with elements of Firth and Wagner’s argument, their general claim that mainstream SLA is lopsidedly cognitivist—and that much of its current practice virtually compels a view of language as radically disconnected from its natural environment, even in studies that seem to investigate language use in that environment—is, in my opinion, all too accurate.11 Taking the just-mentioned critiques of SLA as part of a larger critical endeavor, I would like to delineate my own understanding of some key concepts in the field, and indicate where I find them at odds with a sociocognitive perspective.

Input As dramatized at the beginning of this paper, I see the notion of input in mainstream SLA as coming fairly close to that of a switch or trigger, whether one adopts a Universal Grammar perspective or not. At least one can say that input is of interest in this approach mainly as a stimulus activating an autonomous cognitive learning apparatus, which is assumed to perform certain (often unclearly-specified) processing operations. In this way, the stimulus is converted into some kind of internal grammatical representation, which finds, over time, its proper relationship with other such representations in a rule-governed structuralist system that is virtually immune to nonlinguistic influence. Even SLA studies of input modification and interaction (see below), which focus nominally on language-inuse, seem to adopt this scenario—or something quite like it—as their underlying “central dogma” (Crick, 1988). Much of this characterization is problematic from a sociocognitive perspective. If knowledge of the world, including linguistic knowledge, is organized in the form of “actional wholes” (Hanks, 1996, p. 245, where actional refers to the fundamental purpose of language to contribute to the carrying out of action in the world), it is hard to imagine how such knowledge would develop via decontextualized internalization. The developing grammatical system as I have described it furthermore lacks any motivation other than perhaps a purely genetic one, which is why its description is often pervaded by an odd sort of anomie, in my opinion—it just is because it is. As I have tried to make clear, any approach to language and its acquisition which ignores or dismisses the basic functionality of language for human beings in society is lacking in descriptive and theoretical adequacy.

Interaction Most SLA researchers who study “interaction” do so mainly for the sake of understanding its conditioning effect on input. This is made clear by Gass (1998): The goal of my work (and the work of others within the input/interaction framework . . .) has never been to understand language use per se . . . but rather to understand what types of interaction might bring about what types of changes in linguistic knowledge. (p. 84)

This is truly a pale reflection of the study of authentic human (linguistic) interaction, as Liddicoat (1997) noted: What is missing . . . [in] the study of interaction in L2 contexts is interaction between people who have a preexisting relationship, who are interacting for the purpose of engaging in that relationship, and who are engaged in interaction in which their options for participating are not constrained by institutional roles.12 (p. 314)

In other words, what is missing in mainstream SLA is any concern whatsoever for the dominant forms of interaction in the world, and for those interactions qua interactions. The term therefore seems to be a misnomer, given the fact that the focus of work that purportedly studies it is still language-in-the-head. Again in the words of Liddicoat: “Essentially, what is happening here [in the contrived interactional settings favored in mainstream SLA] is an interaction designed to allow the NNS to produce a language sample [italics added]” (p. 315). From a sociocognitive perspective this is hardly interaction at all. The Language Learner To put it bluntly, the language learner in mainstream SLA is something like an automaton, interesting only in the sense that it houses a discrete language learning system. This view is well attested in writing on the goals of SLA; to restate the words of Long (1997), cited at the beginning of this article: “Most SLA researchers view the object of inquiry as in large part an internal, mental process” (p. 319). If SLA is about language-learning human beings, they are therefore human only in a derivative sense—analogous, more or less, to the attenuated manner in which language is considered “creative” by Chomskyans. As noted previously, I frequently find the reading of SLA research to be almost an exercise in surrealism—based, I believe, in the just-mentioned contradictory “present absence” of human

536 beings. Human beings as I know them, whether people on the street, students I teach and work with, professional colleagues, or those I am close to and love, appear to act, think, and feel in ways and for reasons entirely different than those most typically featured in SLA research. Case studies, diary studies, and (mostly earlier) work that focuses on authentic interaction and language use (e.g., L. W. Fillmore, 1979; Saville-Troike, 1988; Schmidt, 1983; Schmidt & Frota, 1986) are some of the exceptions I know of in this regard—and they are admirable exceptions.

The Modern Language Journal 86 (2002) Cognition In mainstream SLA cognition is a “lonely” process taking place within an autonomous language learning organism. Its forte is the processing of input/construction of linguistic knowledge. It serves as a bank of internal linguistic knowledge, or competence, which most often has only an indirect connection to language performance. Even non-mainstream areas of SLA that focus on language performance, such as interlanguage pragmatics, sometimes seem to assume this view (e.g., Kasper, 1997).

(Scientific) Methods From the start, SLA was “scientized” by its practitioners (Block, 1996). This was partly because it adopted cognitive psychology and linguistics as its source disciplines (e.g., Crookes, 1997; Mitchell & Myles, 1998). As time went on, however, a second good reason for seeing SLA as science became apparent: Academic politics and prestige favored doing so, especially if SLA was thereby dissociated from the low-status occupation of language teaching (Schachter, 1993). Although this move may have brought a certain amount of respect-by-(dis?)association to the field, recent charges of “science envy” (Block, 1996) or “physics envy” (Lantolf, 1996) are, in some respects, not far off the mark. The idea that one can study inherently social phenomena by applying methods and approaches originally developed to describe the behavior of inanimate objects has been widely critiqued, starting as far back, perhaps, as Wundt (see quotation from Breen, 1985, given above). Strong arguments have also been made to the effect that quantitative scientific methodologies, because they neutralize by design what is variable and individual (in human behavior or otherwise), produce epiphenomenally uniform accounts (e.g., Daston, 1995; Hacking, 1981; Porter, 1995). That is, uniform results have been argued to be substantially an artifact of methodologies which, in a sense, presuppose them—for example, by statistical generalization or strict control of “extraneous” variables. Although such critiques have been widespread in the social sciences over the past 20 or so years, it is of more than passing interest that defenders of mainstream scientized SLA (e.g., most of the contributors to Beretta, 1993) tend to rely for their arguments more on the (nonempirical) philosophy of science, than on actual, empirical accounts of scientific behavior (e.g., Collins, 1985; Gilbert & Mulkay, 1984; Latour & Woolgar, 1986; Traweek, 1988).

Language Where language is reduced substantially to grammar, and its use largely to the provision of input, there exists a reductiveness approaching that of Chomksy’s influential vision. From a sociocognitive perspective, however, language is an abundantly rich resource for getting on in the world—for performing social action. Language is intricately but dynamically interwoven with humans’ other means of ecological adaptation and activity, and removing it from that context comes at a real cost. Beyond simply defining a sociocognitive approach to SLA negatively by criticizing SLA orthodoxy, however, I would also like to provide a view of what it is, or, more accurately, what it could be. One thing is clear: Such a perspective does not yet exist in SLA. For this reason, it will be possible to go just so far in conjuring up its image. First, a sociocognitive approach to SLA would take the social dimensions of language and its acquisition seriously. Interaction in sociocognitive SLA would have its full sociocognitive significance and constitute a foundational concept. Language is learned in interaction, often with more capable social members. Classroom teachers are part of this group where second languages are concerned, but only a part—peers, mentors, role models, friends, family members, and significant others can also fall into this category. Although interaction might not include conversation in all cases, it would certainly entail the deep, holistic investment of learners in learning activities, and would see those learners as active agents, not passive recipients. Second, language and its acquisition would be fully integrated into other activities, people, and things in a sociocognitive approach to SLA. They would be seen as integral parts of larger sociocognitive wholes, or, in Gee’s (1992) term, Discourses: “Discourses are composed of people, of ob-

Dwight Atkinson jects . . . and of . . . talking, acting, interacting, thinking, believing, and valuing, and sometimes . . . writing, reading, and/or interpreting . . . . Discourses are out in the world, like books, maps, and cities” (p. 20). In other words, language would be seen in terms of its rich ecological/contextual/relational “worldliness” (Pennycook, 1994) and complexity rather than its simplicity, parsimony, and autonomy. Third, language and its acquisition, from a sociocognitive perspective, would be seen in terms of “action” and “participation”—as providing an extremely powerful semiotic means of performing and participating in activity-in-the-world (Rogoff, 1990, 1998; Lave & Wenger, 1991). It is just because language serves this vital function, one could say, that it exists and that people acquire it: One does not usually acquire a language in order to acquire it, or talk about it, or provide data for SLA researchers. One acquires a language in order to act, and by acting, in a world where language is performative. This is exactly why and how children learn their first language, and it accounts as well for most of the second/additional language learning going on in the world today. It is an extremely unfortunate matter that we have some influential myths suggesting otherwise—for example, that SLA takes place primarily in cases where students are taught something discrete called language (Gee, 1990), or that SLA takes place within isolated heads whose only motivation is creative construction. Finally, a sociocognitive perspective should not, strictly speaking, exclude. As an approach to language, it is fundamentally cognitive and fundamentally social. A sociocognitive perspective does not diminish a view of language as one or the other—it argues for the profound interdependency and integration of both. Thus, although I have described a cognitivist view of SLA as “impoverished,” that is not the same as saying it is wrong. In fact, I believe that cognitivists have contributed to our understanding—however tentative and partial at this point—of SLA. But if language is in the world at the same time as it is in the head, then we need to account for its integrated existence, rather than adopt positions that reduce the life—the humanity—out of language. A NOTE ON NEO-VYGOTSKIAN SOCIOCULTURAL THEORY IN SLA Given the increasing popularity of a non-mainstream form of SLA theory—neo-Vygotskian sociocultural theory—which tries to address some

537 of the same problems with mainstream SLA research discussed in this article, I would like here briefly to indicate where I believe my approach differs from this one and is not simply reducible to it. A major claim of Vygotsky—and a claim that seems typically to be assumed or endorsed by current neo-Vygotskian SLA researchers (e.g., Lantolf, 2000)—was that language appears, developmentally speaking, first on the interpsychological plane (i.e., as social speech) and only afterwards on the intrapsychological plane (i.e., as internalized or inner speech). There is thus a gradual process of internalization whereby a fully externalized social practice becomes a substantially internalized cognitive practice (e.g., Lantolf, 2000; Lantolf with Pavlenko, 1995). In fact, according to Vygotsky, this is how people learn to cognize—how “higher mental development” occurs: “Any higher mental function was external because it was social at some point before becoming an internal, truly mental function [italics added]” (Vygotsky, 1981, p. 162). In a recent major statement of sociocultural theory in SLA, Lantolf (2000) has echoed Vygotsky’s view: Internalization is in essence the process through which higher forms of mentation come to be. Internalization then assumes that the [original] source of consciousness resides outside of the head and is in fact anchored in social activity. At first the activity of individuals is organized and regulated (i.e., mediated) by others, but eventually, in normal development, we come to organize and regulate our own mental and physical activity through the appropriation of the regulatory means employed by others. At this point psychological functioning comes under the voluntary control of the person. (pp. 13–14)

From this point of view, we could say that language is sociocognitive in a manner of speaking, but only or most substantially in its developmental profile (see also Kirshner & Whitson, 1997, for a similar critique). That is, language starts social, but becomes substantially cognitive as development proceeds: in the words, once again, of Lantolf (2000), “The convergence of thinking [i.e., the cognitive] with culturally created mediational artifacts [i.e., the social] . . . occurs in the process of internalization, or the reconstruction on the inner, psychological plane, of socially mediated external forms of goal-directed activity” (p. 13). For some neo-Vygotskians this may be a sufficient explanation of both language development and use in toto, given that for Vygotsky, “to understand behavior means to understand the history of

538 behavior [italics added]”(cited in Cole, 1993, p. 9).13 But my own argument—as developed on a stepby-step basis in this article—is that although language may perhaps be seen from some points of view as more or less internalized and self-regulated—as the property of an individual, cognitive self—in actuality it always and everywhere exists in an integrated sociocognitive space. In terms of the above-given formulation of Vygotsky (1981), I would therefore argue that, contra his view, language (no matter what its stage of development) never takes on an “internal, truly mental function” at all. Rather, it is always mutually, simultaneously, and co-constitutively in the head and in the world. Certainly, at critical points in development, language may seem to “come inside”—but if one end of language, so to speak, is embedded in cognitive space, the other end is just as strongly embedded in social space. In actual fact, this use of spatial and object-oriented metaphors to describe language hardly suffices, but I hope at least to have indicated what I believe is a basic difference—certainly at least in terms of emphasis—between what I propose here and what many neo-Vygotkians assume or are proposing in their SLA work. There are other differences—for instance, the artificial tasks and conditions that have historically obtained in Vygotskian research, and the need in such research to break behavior down into component parts while at the same time professing holism (e.g., Wertsch, 1998; see also Kirshner & Whitson, 1997). But the first-mentioned difference is the most important one, and from it many of the other differences proceed. This is not to say that I do not value or respect research being done in this framework in the field, nor that it has not contributed significantly to my own thinking on SLA.

CONCLUSION What, then, are some of the implications of a sociocognitive view of SLA? Here I will speculate because, as I have already pointed out, no such coherent view yet exists. In the interim, I will take my cues largely from pedagogically-oriented L1 research on situated cognition, social linguistics, and learning-as-participation (e.g., Gee, 1990, 1992; Kirshner & Whitson, 1997; Lave & Wenger, 1991; Rogoff, 1990, 1998; Wenger, 1998; cf. Atkinson, 1997). I should also note that although several of the implications described below are in areas that are already receiving attention in sec-

The Modern Language Journal 86 (2002) ond language education, they have not yet in most cases been closely linked to SLA theory. One implication of a sociocognitive approach to SLA is that teaching is valuable, and that learning and teaching go hand-in-hand (a view that mainstream SLA, incidently, has yet to agree on—Crookes, 1997; R. Ellis, 1997). But teaching in this view is also often incidental: If one learns by participating in specific and meaningful social activity, then co-participants are often one’s teachers. The expert-novice (or master-apprentice) metaphor emanating from studies of situated cognition is a useful thinking tool in this context, exactly because it emphasizes learning through active and increasingly knowledgeable participation in a particular “community of practice.” But this metaphor should not be taken to imply that learners and those they learn from are profoundly separated. As already noted, peers can be teachers, depending on the situation, and the same is true for all others with whom we have social relationships. If we consider that most of the additional language learning going on in the world today is of this nonformal variety, then we have a more realistic notion of teaching, and how teaching and learning interact. If, as second language teachers, we can harness more of the range of teaching situations that actually take place in the world outside the classroom (e.g., Atkinson, 1997, 1998; Hawkins, 1998), then we will be able to utilize more fully the teaching and learning potential of all human beings. A second implication of a sociocognitive approach to SLA is intimately related to the view that language and its acquisition are not radically disconnected from the rest of the world. Language-in-the-world suggests a richness and power for it that extends well beyond the transfer of information from brain to brain. A sociocognitive approach to SLA promotes and reinforces many connections to other realms of inquiry and practice, such as: culture (e.g., Atkinson, 1999; Kramsch, 1993); schooling (e.g., Poole, 1992); identity (e.g., Peirce, 1995); power, politics, and ideology (e.g., Fairclough, 1992; Gee, 1990, 1992; Pennycook, 1994); discourse (e.g., Firth & Wagner, 1997); social ecology (e.g., Capra, 1996; Schumacher, 1997); and embodied action-taking (e.g., Goodwin, 2000). When one considers the deep involvement of second/additional languages in world politics, exploitative capitalism, and globalization (e.g., Pennycook, 1994, 1998; Phillipson, 1992), then a very real gain is realized in the intrinsic consequentiality of studying, learning, and teaching them. We need and can have a SLA—not to mention approaches to sec-

539

Dwight Atkinson ond language teaching—with real potential for changing the world, rather than being radically separated from it. A third implication of the sociocognitive study of SLA concerns research methods for studying the learning and teaching of second/additional languages. Although individual-aggregating statistical and experimental methods have a place in SLA, studying real humans in real human contexts and interactions, including classrooms, entails methodologies that do not denature phenomena by removing them from their natural environments and breaking them down into countable component parts. Qualitative research approaches that attempt to honor the profound wholeness and situatedness of social scenes and individuals-in-the-world, such as ethnographic methods (e.g., Holliday, 1996; Lazaraton & Davis, 1995; Ramanathan & Atkinson, 1999), will have a central place in this endeavor. One further important advantage of such methodologies is that they are less likely to reproduce the theory-practice divide—the asymmetrical division of labor between (classroom) teachers and researchers. In the suggestive phrasing of Holliday (1994), “the teacher cannot afford to be anything but a researcher” (p. 31) where broad sociocognitive connections are acknowledged between the classroom and the outside world. Likewise, classrooms, among other kinds of learning situations, can be studied via these methods as complex sociocognitive activity systems in their own right (Breen, 1985; Holliday, 1996). Fourth and finally, sociocognitive approaches to SLA will provide a means by which second language learners can be seen as real people, doing something they naturally do—not as mere research subjects, or mere students, or mere sites for language acquisition. Perhaps the theme I have emphasized above others in this article is that thought, feeling, and activity in the social world are brought together in the form of human beings actively operating as part of that world. It is therefore fitting to give the last word to Lave and Wenger (1991), two scholars who have done much to promote this view: Participation in social practice . . . suggests a very explicit focus on the person, but person-in-the-world, as member of a sociocultural community. This focus in turn promotes a view of knowing as activity by specific people in specific circumstances. As an aspect of social practice, learning involves the whole person: it implies not only a relation to specific activities, but a relation to social communities—it implies becoming a full participant, a member, a kind of person. . . . Activities, tasks, functions, and under-

standings do not exist in isolation; they are part of broader systems of relations in which they have meaning. These systems of relations arise out of and are reproduced and developed within social communities, which are in part systems of relations among persons. The person is defined by as well as defines these relations. Learning thus implies becoming a different person with respect to the possibilities enabled by these systems of relations. To ignore this aspect of learning is to overlook the fact that learning involves the construction of identities. (pp. 52–53)

ACKNOWLEDGMENTS

James Paul Gee first introduced me to many of the ideas developed in this article. I would like to thank him, as well as A. Vishnu Bhat, Michelle Burnham, Joan Carson, John Hedgcock, Hiroko Itakura, Robert B. Kaplan, Wakako Kobayashi, Claire Kramsch, Ilona Leki, Takako Nishino, and Nirmal Selvamony, for their critical comments and encouragement in what has been a drawnout writing/publishing process. David Beglar, Suresh Canagarajah, George Kamberelis, Genevieve Patthey-Chavez, and Vai Ramanathan also generously read and commented on earlier drafts. I would further like to thank the MLJ editor, Sally Sieloff Magnan, and the four anonymous reviewers for their thoughtful editing and comments. This article is dedicated to the memory of Matha Bean. NOTES 1 The sociocognitive approach developed here takes its lead from a wide variety of disciplines and approaches: cognitive and cultural anthropology (e.g., Hanks, 1996; Shore, 1996; Strauss & Quinn, 1997); sociology (e.g., Berger & Luckmann, 1966; Goffman, 1959; Sacks et al., 1974); social linguistics (Gee, 1990, 1992); sociolinguistics and language socialization studies (e.g., Hymes, 1972; Ochs, 1988a, 1988b, 1990, 1992, 1996; Schieffelin & Ochs, 1986); neo-Vygotskian sociocultural theory (e.g., Lantolf, 2000; Lantolf & Appel, 1994; Vygotsky, 1981; Wertsch, 1985, 1998); studies of situated cognition/learning and communities of practice (e.g., Brown, Collins, & Duguid, 1989; Kirshner & Whitson, 1997; Lave & Wenger, 1991; Rogoff, 1990; Wenger, 1998); cognitive science and connectionism (e.g., Churchland, 1996; A. Clark, 1997; Elman et al., 1996; Rumelhart & McClellan, 1986; Seidenberg, 1996); L1 acquisition (e.g., Bates et al., 1995; Foster, 1990; Plunkett, 1995); SLA (e.g., Breen, 1985; Firth & Wagner, 1997; Peirce, 1995); and studies of conversation and interaction (e.g., Goodwin, 1986, 1987, 2000; Lerner, 1993, 1996; Ochs et al., 1996). The term sociocognitive has been used in a variety of

540 contexts in a number of different disciplines, some of which are listed above. My own first exposure to the term came in a graduate class in sociolinguistics taught by Elinor Ochs, one of the main proponents of some of the views I develop here (e.g., Ochs, 1988a; Ochs et al., 1996). 2 As this is a historical description, I do not refer to more recent Chomskyan formalisms. In fact, however, Chomsky’s basic theory of language (as opposed to its ever-changing formal realizations) has remained remarkably stable over the years (e.g., Chomsky, 2000). 3 This same basic point is at least implicit in Hymes’s (1972) original conceptualization of communicative competence, which I regard as foundational to a sociocognitive view of language. 4 The point has even been made (e.g., Goffman, 1959) that, given an utterance in which the propositional content and the nonpropositional content are in conflict, the latter is usually taken as the final arbiter of meaning. 5 Implicit in most definitions of connectionism is the notion of parallel distributed processing—often taken to be its virtual synonym. Parallel (vs. serial) processing involves the simultaneous carrying out of large numbers of computational events (vs. the step-by-step nature of serial processing), whereas distributed (versus central) processing indicates that cognitive computational activity is spread out over a large number of neurons/neural connections, rather than having to depend on any one or a small number of them. 6 This approach is part of a broader critique across the social sciences of mainstream psychology as the study of “lonely cognition,” instead of self-in-context (e.g., Gergen, 1985; Hutchins, 1995; Lave & Wenger, 1991; Rogoff, 1990, 1998). A deeper and more philosophical critique of the self is a major theme in postmodernist and poststructuralist thought (see, e.g., Atkinson, 1999) but is beyond the scope of this article. 7 At the same time, participants hew closely to the conventional story structure and register of narratives, or if they do not are usually brought into line. 8 Deixis is perhaps the paradigm example here, but the same basic claim can be made across linguistic domains and systems (Hanks, 1996; Ochs, 1992). 9 In general, the descriptions given in this section are incomplete as they stand in that they do not sufficiently capture the bidirectionality of influence and interpenetration of social and cognitive processes. Because I am partly focusing on cognition here—specifically as it concerns the issue of “decontextualized language”—I have necessarily had to downplay the reciprocal influence of the cognitive on the social, as well as the continuous and integrated nature of feedback between them. To give two examples by way of atonement and the restoration of holistic bidirectionality: (a) The design of our cognitive apparatus fundamentally influences the way we perceive things, so that social practices and products must be preadapted to our cognitive-sensory capabilities; and (b) Gee’s (1992) description of the sociocognitive schemas of cities demonstrates profound sociocognitive interpenetration and interaction. In fact, although I will

The Modern Language Journal 86 (2002) not seek to articulate them here, substantial areas of the interpenetration of cognitive and social phenomena must operate much more dynamically and dialectically—perhaps along the lines of Giddens’ (1979) notion of structuration (see also Atkinson, 1999). Thus, it can be said that individuals and social groups not only progressively adapt to their environments, but that they also actively construct those environments in many and varied ways. Larsen-Freeman’s (1997) third meaning of the term “dynamic” (p. 148) in describing language from a chaos/complexity perspective—although it still seems to depend in some senses on the much-questioned notion of linguistic rules—captures something of the interactivity, feedback-sensitivity, and constructedness of “languaging” (Larsen-Freeman herself uses the term grammaring) I am trying to get at here. In general, I find Larsen-Freeman’s discussion of chaos/complexity theory vis-à-vis SLA inspirational in its potential for refocusing the field. 10 This view seems problematic if one is to define language more broadly than a set of grammatical forms and rules. How would one account for the acquisition of norms of politeness (e.g., Brown & Levinson, 1987; Tannen, 1986), for instance, other than by appealing to the direct effect of social factors? Tannen has shown in various publications that politeness is an aspect of language that is central to its use and functionality—every bit as important as the grammatical features more standardly studied in SLA. 11 Two SLA researchers whose work has been largely within the cognitivist tradition seem recently to have realized the reductionism inherent in that tradition. R. Ellis (1997)—partly cited at the beginning of this article—and Crookes (1997), respectively, state: “SLA in general has paid little attention to the social context of L2 acquisition, particularly where context is viewed nondeterministically (i.e., as something learners construct for themselves). SLA has been essentially a psycholinguistic enterprise, dominated by the computational metaphor of acquisition” (R. Ellis, 1997, p. 87). “Though cognitive psychology [as a source discipline for SLA research] was to be preferred to its dominant predecessor [i.e., behaviorism] because it was (purportedly, at least) about people (rather than rats), it was a long time before I began to understand that it . . . could be seen as a sociocultural construct . . . that reflected at least to some extent the presumptions of the societies in which it developed. That was why it was fundamentally an individualist psychology that treats people as isolates” (Crookes, 1997, p. 98). It is noteworthy that both of these statements were made in articles discussing the relevance of SLA to second language teaching. 12 Although any claim that participation in interaction between familiars is not constrained by institutional roles is highly questionable (e.g., Bourdieu, 1984), Liddicoat is clearly making a distinction here between what conversation analysts call “ordinary conversation” and what they call “institutional interaction” (for more on this distinction, see Drew & Heritage, 1992). 13 If this is not an understanding of language shared

Dwight Atkinson by all neo-Vygotskians (e.g., Rogoff, 1998; Wertsch, 1998), it still seems a dominant emphasis among neo-Vygotskians who study SLA, as seen in the Lantolf quotations given earlier in this section. To me, this emphasis represents a particularly literalistic interpretation of Vygotsky’s original thought.

REFERENCES Atkinson, D. (1997). A critical approach to critical thinking in TESOL. TESOL Quarterly, 31, 71–94. Atkinson, D. (1998). Forum response to comments on “A critical approach to critical thinking in TESOL.” TESOL Quarterly, 32, 133–137. Atkinson, D. (1999). TESOL and culture. TESOL Quarterly, 33, 625–654. Bakhtin, M. M. (1990). Marxism and the philosophy of language. In P. Bizzell & B. Herzberg (Eds.), The rhetorical tradition: Readings from classical times to the present (pp. 928–944). New York: St. Martins. Barrett, M. (1995). Early lexical development. In P. Fletcher & B. MacWhinney (Eds.), The handbook of child language (pp. 362–392). Oxford, UK: Blackwell. Bates, E., Dale, P. S., & Thal, D. (1995). Individual differences and their implications for theories of language development. In P. Fletcher & B. MacWhinney (Eds.), The handbook of child language (pp. 96–151). Oxford, UK: Blackwell. Becker, A. L. (1979). Text-building, epistemology, and aesthetics in Javanese shadow theatre. In A. L. Becker & A. A. Yengoyan (Eds.), The imagination of reality: Essays in southeast Asian coherence systems (pp. 211–243). Norwood, NJ: Ablex. Becker, A. L. (1988). Language in particular: A lecture. In D. Tannen (Ed.), Linguistics in context: Connecting observation and understanding. Advances in discourse processes (Vol. 29, pp. 17–35). Norwood, NJ: Ablex. Beretta, A. (Ed.). (1993). Theory construction in SLA [Special issue]. Applied Linguistics, 14(3). Berger, P. L., & Luckmann, T. (1966). The social construction of reality. New York: Doubleday. Block, D. (1996). Not so fast: Some thoughts on theory culling, relativism, accepted findings and the heart and soul of SLA. Applied Linguistics, 17, 63–83. Bourdieu, P. (1984). Distinction: A social critique of the judgement of taste. Cambridge, MA: Harvard University Press. Bourdieu, P. (1991). Language and symbolic power. Cambridge, MA: Harvard University Press. Bower, T. G. R., & Wishart, J. G. (1979). Towards a unitary theory of development. In E. B. Thoman (Ed.), Origins of the infant’s social responsiveness (pp. 65–93). Hillsdale, NJ: Erlbaum. Breen, M. P. (1985). The social context of language learning—A neglected situation? Studies in Second Language Acquisition, 7, 135–158. Brown, J. S., Collins, A., & Duguid, P. (1989). Situated

541 cognition and the culture of learning. Educational Researcher, 18, 32–42. Brown, P., & Levinson, S. C. (1987). Politeness: Some universals of language usage. Cambridge: Cambridge University Press. Brown, R. (1973). A first language: The early stages. Cambridge, MA: Harvard University Press. Capra, F. (1996). The web of life: A new scientific understanding of living systems. New York: Doubleday. Chomsky, N. (2000). New horizons in the study of language and mind. Cambridge: Cambridge University Press. Churchland, P. M. (1996). The engine of reason, the seat of the soul: A philosophical journey into the brain. Cambridge, MA: MIT Press. Clark, A. (1997). Being there: Putting brain, body, and world together again. Cambridge, MA: MIT Press. Clark, E. V. (1983). Meanings and concepts. In P. Mussen (Series Ed.) & J. H. Flavell & E. M. Markman (Vol. Eds.), Handbook of child psychology: Vol. 3. Cognitive development (4th ed., pp. 687–782). New York: Wiley. Clark, H. H., & Marshall, C. R. (1981). Definite reference and mutual knowledge. In A. K. Joshi, B. Webber, & I. Sag (Eds.), Elements of discourse understanding (pp. 10–63). Cambridge: Cambridge University Press. Clark, H. H., & Wilkes-Gibbs, D. (1986). Referring as a collaborative process. Cognition, 22, 1–39. Cole, M. (1985). The zone of proximal development: Where culture and cognition create each other. In J. Wertsch (Ed.), Culture, communication and cognition: Vygotskian perspectives (pp. 146–161). Cambridge: Cambridge University Press. Cole, M. (1993, April). Vygotsky and writing: Reflections on a distant discipline. Paper presented at Conference on College Composition and Communication, San Diego, CA. Collins, H. M. (1985). Changing order: Replication and induction in scientific practice. London: Sage. Corder, H. P. (1967). The significance of learners’ errors. International Review of Applied Linguistics, 5, 161–169. Crick, F. (1988). What mad pursuit: A personal view of scientific discovery. New York: Basic Books. Crookes, G. (1997). SLA and language pedagogy: A socioeducational perspective. Studies in Second Language Acquisition, 19, 93–116. Daston, L. (1995). The moral economy of science. Osiris, 10, 3–24. Davis, K. (1995). Qualitative theory and methods in applied linguistic research. TESOL Quarterly, 29, 427–454. Drew, P., & Heritage, J. (1992). Analyzing talk at work: An introduction. In P. Drew & J. Heritage (Eds.), Talk at work: Interaction in institutional settings (pp. 3–65). Cambridge: Cambridge University Press. Edelsky, C. (1981). Who’s got the floor? Language in Society, 10, 383–421. Eimas, P. D. (1975). Speech perception in early infancy. In L. B. Cohen & P. Salapatek (Eds.), Infant percep-

542 tion: From sensation to cognition (Vol. 2, pp. 193–231). New York: Academic Press. Ellis, N. C. (1998). Emergentism, connectionism, and language learning. Language Learning, 48, 631–664. Ellis, R. (1994). The study of second language acquisition. Oxford: Oxford University Press. Ellis, R. (1997). SLA and language pedagogy: An educational perspective. Studies in Second Language Acquisition, 19, 69–92. Elman, J. L. (1992). Learning and development in neural networks: The importance of starting small. In C. Umilta & M. Moscovitch (Eds.), Attention and performance IX: Conscious and nonconscious information processing (pp. 861–889). Hillsdale, NJ: Erlbaum. Elman, J. L., Bates, E. A., Johnson, M. H., KarmiloffSmith, A., Parisi, D., & Plunkett, K. (1996). Rethinking innateness: A connectionist perspective on development. Cambridge, MA: MIT Press. Fairclough, N. (1992). Discourse and social change. Cambridge, UK: Polity. Fillmore, C. J. (1975). An alternative to checklist theories of meaning. In Proceedings of the First Annual Meeting of the Berkeley Linguistics Society (pp. 123–131). Berkeley, CA: University of California. Fillmore, C. J. (1985). Frames and the semantics of understanding. Quaderni di Semantica, 6, 222–253. Fillmore, C. J., Kay, P., & O’Connor, M. C. (1988). Regularity and idiomaticity in grammatical constructions. Language, 64, 501–538. Fillmore, L. W. (1979). Individual differences in second language acquisition. In C. J. Fillmore (Ed.), Individual differences in language ability and language behavior (pp. 203–228). San Diego, CA: Academic Press. Firth, A., & Wagner, J. (1997). On discourse, communication, and (some) fundamental concepts in SLA research. Modern Language Journal, 81, 285–300. Foster, S. H. (1990). The communicative competence of young children. London: Longman. Gass, S. (1998). Apples and oranges: Or, why apples are not orange and don’t need to be: A response to Firth and Wagner. Modern Language Journal, 82, 83–90. Gee, J. P. (1990). Social linguistics and literacies: Ideology in discourses. London: Falmer Press. Gee, J. P. (1992). The social mind. London: Bergin & Garvey. Gee, J. P. (1995). First language acquisition as a guide for theories of learning and pedagogy. Linguistics and Education, 6, 331–354. Gergen, K. (1985). The social constructionist movement in modern psychology. American Psychologist, 40, 266–275. Giddens, A. (1979). Central problems in social theory: Action, structure and contradiction in social analysis. Berkeley, CA: University of California Press. Gilbert, G. N., & Mulkay, M. (1984). Opening Pandora’s

The Modern Language Journal 86 (2002) box: A sociological analysis of scientific discourse. Cambridge: Cambridge University Press. Goffman, E. (1959). The presentation of self in everyday life. New York: Anchor/Doubleday. Goodwin, C. (1981). Conversational organization: Interaction between speakers and hearers. New York: Academic Press. Goodwin, C. (1986). Audience diversity, participation and interpretation. Text, 6, 283–316. Goodwin, C. (1987). Unilateral departure. In G. Button & J. R. E. Lee (Eds.), Talk and social organization (pp. 206–216). Clevedon, UK: Multilingual Matters. Goodwin, C. (2000). Action and embodiment within situated human interaction. Journal of Pragmatics, 32, 1489–1522. Goodwin, C., & Goodwin, M. H. (1992). Assessments and the construction of context. In A. Duranti & C. Goodwin (Eds.), Rethinking context: Language as an interactive phenomenon (pp. 147–189). Cambridge: Cambridge University Press. Gould, S. J. (1981). The mismeasure of man. New York: Norton. Gregg, K. (1988). Second language acquisition theory: The case for a generative perspective. In S. M. Gass & J. Schachter (Eds.), Linguistic perspectives on second language acquisition (pp. 15–40). New York: Cambridge University Press. Gruendel, J. (1977). Referential extension in early language development. Child Language, 48, 567–576. Gumperz, J. J. (1982). Discourse strategies. Cambridge: Cambridge University Press. Hacking, I. (1981). Representing and intervening. Cambridge: Cambridge University Press. Hall, J. K. (1997). A consideration of SLA as a theory of practice: A response to Firth and Wagner. Modern Language Journal, 81, 301–306. Halliday, M. A. K. (1978). Language as social semiotic: The social interpretation of language and meaning. London: Edward Arnold. Halliday, M. A. K., & Hasan, R. (1989). Language, context, and text: Aspects of language in a social-semiotic perspective. Oxford: Oxford University Press. Hanks, W. F. (1996). Language form and communicative practices. In J. J. Gumperz & S. C. Levinson (Eds.), Rethinking linguistic relativity (pp. 232–270). Cambridge: Cambridge University Press. Hawkins, M. (1998). Forum response to Dwight Atkinson’s “A critical approach to critical thinking in TESOL.” TESOL Quarterly, 32, 127–132. Holland, D., & Quinn, N. (1987). Cultural models in language and thought. Cambridge: Cambridge University Press. Holliday, A. (1994). Appropriate methodology in social context. Cambridge: Cambridge University Press. Holliday, A. (1996). Developing a sociological imagination: Expanding ethnography in international English language education. Applied Linguistics, 17, 234–255. Hopper, P. (1988). Emergent grammar and the a priori grammatical postulate. In D. Tannen (Ed.), Lin-

Dwight Atkinson guistics in context: Connecting observation and understanding. Advances in discourse processes (Vol. 29, pp. 117–134). Norwood, NJ: Ablex. Hutchins, E. (1995). Cognition in the wild. Cambridge, MA: MIT Press. Hymes, D. (1972). On communicative competence. In J. B. Pride & J. Holmes (Eds.), Sociolinguistics: Selected readings (pp. 269–293). Harmondsworth, UK: Penguin. Kasper, G. (1997). “A” stands for acquisition: A response to Firth and Wagner. Modern Language Journal, 81, 307–312. Kemper, S. (1987). Life-span changes in syntactic complexity. Journal of Gerontology, 42, 323–328. Kendon, A. (1992). The negotiation of context in faceto-face interaction. In A. Duranti & C. Goodwin (Eds.), Rethinking context: Language as an interactive phenomenon (pp. 323–334). Cambridge: Cambridge University Press. Kirshner, D., & Whitson, J. A. (Eds.). (1997). Situated cognition theory: Social, neurological, and semiotic perspectives. Mahwah, NJ: Lawrence Erlbaum. Kramsch, C. (1993). Context and culture in language teaching. Oxford: Oxford University Press. Langacker, R. W. (1987). Foundations of cognitive grammar, Vol. 1. Stanford, CA: Stanford University Press. Lantolf, J. P. (1996). SLA theory building: “Letting all the flowers bloom!” Language Learning, 46, 713–749. Lantolf, J. P. (Ed.). (2000). Sociocultural theory and second language learning. Oxford: Oxford University Press. Lantolf, J. P., & Appel, G. (1994). Vygotskian approaches to second language research. Norwood, NJ: Ablex. Lantolf, J. P. (with Pavlenko, A.). (1995). Sociocultural theory and second language acquisition. Annual Review of Applied Linguistics, 15, 108–124. Larsen-Freeman, D. (1991). Second language acquisition research: Staking out the territory. TESOL Quarterly, 25, 315–350. Larsen-Freeman, D. (1997). Chaos/complexity science and second language acquisition. Applied Linguistics, 18, 141–165. Latour, B., & Woolgar, S. (1986). Laboratory life: The construction of scientific facts (2nd ed.). Princeton, NJ: Princeton University Press. Lave, J., & Wenger, E. (1991). Situated learning: Legitimate peripheral participation. Cambridge: Cambridge University Press. Lazaraton, A., & Davis, K. (Eds.). (1995). Qualitative research in ESOL [Special issue]. TESOL Quarterly, 29 (4). Leonard, L. B. (1995). Phonological impairment. In P. Fletcher & B. MacWhinney (Eds.), The handbook of child language (pp. 573–602). Oxford, UK: Blackwell. Lerner, G. (1993). Collectivities in action: Establishing the relevance of conjoined participation in conversation. Text, 13, 213–245. Lerner, G. (1996). On the “semi-permeable” character

543 of grammatical units in conversation: Conditional entry into the turn space of another speaker. In E. Ochs, E. A. Schegloff, & S. A. Thompson (Eds.), Interaction and grammar (pp. 238–276). Cambridge: Cambridge University Press. Liddicoat, A. (1997). Interaction, social structure, and second language use: A response to Firth and Wagner. Modern Language Journal, 81, 313–317. Locke, J. L. (1993). The child’s path to spoken language. Cambridge, MA: Harvard University Press. Locke, J. L. (1995). Development of the capacity for spoken language. In P. Fletcher & B. MacWhinney (Eds.), The handbook of child language (pp. 278–302). Oxford, UK: Blackwell. Long, M. L. (1997). Construct validity in SLA research: A response to Firth and Wagner. Modern Language Journal, 81, 318–323. Malinowski, B. (1923). The problem of meaning in primitive languages. In C. A. Ogden & I. A. Richards (Eds.), The meaning of meaning: A study of the influence of language upon thought and of the science of symbolism (pp. 296–336). New York: Harcourt, Brace & World. Mandelbaum, J. (1987). Recipient-driven storytelling in conversation. Unpublished doctoral dissertation, University of Texas, Austin. Marchman, V., & Bates, E. (1994). Continuity in lexical and morphological development: A test of the critical mass hypothesis. Journal of Child Language, 21, 339–366. McDermott, R. (1988). Inarticulateness. In D. Tannen (Ed.), Linguistics in context: Connecting observation and understanding. Advances in discourse processes (Vol. 29, pp. 37–68). Norwood, NJ: Ablex. McLaughlin, B. (1987). Theories of second-language learning. London: Arnold. McNeill, D. (Ed.). (2000). Language and gesture. Cambridge: Cambridge University Press. Mitchell, R., & Myles, F. (1998). Second language learning theories. London: Arnold. Nelson, K. (1973). Structure and strategy in learning to talk [Monograph]. Monographs of the Society for Research in Child Development, 38. Nelson, K. (Ed.). (1986). Event knowledge: Structure and function in development. Hillsdale, NJ: Erlbaum. Newson, J. (1979). The growth of shared understandings between infant and caregiver. In M. Bullowa (Ed.), Before speech: The beginning of interpersonal communication (pp. 207–222). Cambridge: Cambridge University Press. Ninio, A., & Bruner, J. (1978). The achievements and antecedents of labelling. Journal of Child Language, 5, 1–15. Ochs, E. (1988a). Culture and language development: Language acquisition and language socialization in a Samoan village. Cambridge: Cambridge University Press. Ochs, E. (1988b, Fall). Introduction to sociolinguistics [classroom lecture]. Linguistics 510. University of Southern California Linguistics Department. Ochs, E. (1990). Indexicality and socialization. In G.

544 Herdt, R. Shweder, & J. Stigler (Eds.), Cultural psychology: Essays on comparative human development (pp. 287–308). Cambridge: Cambridge University Press. Ochs, E. (1992). Indexing gender. In A. Duranti & C. Goodwin (Eds.), Rethinking context: Language as an interactive phenomenon (pp. 335–358). Cambridge: Cambridge University Press. Ochs, E. (1996). Linguistic resources for socializing humanity. In J. J. Gumperz & S. C. Levinson (Eds.), Rethinking linguistic relativity (pp. 407–437). Cambridge: Cambridge University Press. Ochs, E., Schegloff, E. A., & Thompson, S. A. (Eds.). (1996). Interaction and grammar. Cambridge: Cambridge University Press. Ochs, E., Schieffelin, B. B., & Platt, M. L. (1979). Propositions across utterances and speakers. In E. Ochs & B. B. Schieffelin (Eds.), Developmental pragmatics (pp. 251–268). New York: Academic Press. Peirce, B. N. (1995). Social identity, investment, and language learning. TESOL Quarterly, 29, 9–31. Pennycook, A. (1994). The cultural politics of English as an international language. London: Longman. Pennycook, A. (1997). Cultural alternatives and autonomy. In P. Benson & P. Voller (Eds.), Autonomy and independence in language learning (pp. 35–53). London: Longman. Pennycook, A. (1998). English and the discourses of colonialism. London: Routledge. Philips, S. U. (1983). The invisible culture: Communication in classroom and community on the Warm Springs Indian Reservation. New York: Longman. Phillipson, R. (1992). Linguistic imperialism. Oxford: Oxford University Press. Plunkett, K. (1995). Connectionist approaches to language acquisition. In P. Fletcher & B. MacWhinney (Eds.), The handbook of child language (pp. 36–72). Oxford, UK: Blackwell. Plunkett, K., & Marchman, V. (1993). From rote learning to system building: Acquiring verb morphology in children and connectionist nets. Cognition, 48, 21–69. Poole, D. (1992). Language socialization in the second language classroom. Language Learning, 42, 593–616. Porter, T. M. (1995). Trust in numbers: The pursuit of objectivity in science and public life. Princeton, NJ: Princeton University Press. Quirk, R., Greenbaum, S., Leech, G., & Svartvik, J. (1985). A comprehensive grammar of the English language. London: Longman. Ramanathan, V., & Atkinson, D. (1999). Ethnographic approaches and methods in L2 writing research: A critical guide and review. Applied Linguistics, 20, 44–70. Rogoff, B. (1990). Apprenticeship in thinking: Cognitive development in social context. New York: Oxford University Press. Rogoff, B. (1998). Cognition as a collaborative process. In D. Kuhn & R. S. Siegler (Eds.), Handbook of child

The Modern Language Journal 86 (2002) psychology, 5th ed., Vol 2: Cognition, perception, and language (pp. 679–744). New York: Wiley. Rumelhart, D. E., & McClelland, J. L. (1986). On learning the past tense of English verbs. In J. L. McClelland, D. E. Rumelhart, & the PDP Research Group (Eds.), Parallel distributed processing: Explorations in the microstructure of cognition: Vol. 2: Psychological and biological models (pp. 216–271). Cambridge, MA: MIT Press. Sacks, H., Schegloff, E. A., & Jefferson, G. (1974). A simplest systematics for the organization of turntaking for conversation. Language, 50, 696–735. Saffran, J. R., Aslin, R. N., & Newport, E. L. (1996). Statistical learning by 8-month old infants. Science, 274, 1926–1928. Saville-Troike, M. (1988). Private speech: Evidence for second language learning strategies during the silent period. Journal of Child Language, 15, 567–590. Schachter, J. (1993). SLA: Perceptions and possibilities. Second Language Research, 9, 173–187. Schieffelin, B. B., & Ochs, E. (1986). Language socialization. Annual Review of Anthropology, 15, 163–191. Schmidt, R. W. (1983). Interaction, acculturation, and the acquisition of communicative competence. In N. Wolfson & E. Judd (Eds.), Sociolinguistics and language acquisition (pp. 137–174). Rowley, MA: Newbury House. Schmidt, R. W., & Frota, S. N. (1986). Developing basic conversational ability in a second language: A case study of an adult learner of Portuguese. In D. Day (Ed.), Talking to learn: Conversation in second language acquisition (pp. 237–326). Rowley, MA: Newbury House. Schumacher, E. F. (1997). This I believe, and other essays. Foxhole, UK: Green Books. Scollon, R. (1976). Conversations with a one year old. Honolulu, HI: University of Hawai‘i Press. Scribner, S., & Cole, M. (1981). The psychology of literacy. Cambridge, MA: Harvard University Press. Seidenberg, M. S. (1996). Language acquisition and use: Learning and applying probabilistic constraints. Science, 275, 1599–1603. Shore, B. (1996). Culture in mind: Cognition, culture, and the problem of meaning. New York: Oxford University Press. Smith, F. (1988). Joining the literacy club: Further essays into education. Portsmouth, NH: Heinemann. Snow, C. E. (1995). Issues in the study of input: Finetuning, universality, individual and developmental differences, and necessary causes. In P. Fletcher & B. MacWhinney (Eds.), The handbook of child language (pp. 180–193). Oxford, UK: Blackwell. Stemberger, J. (1992). A connectionist view of child phonology: Phonological processing without phonological processes. In C. Ferguson, L. Menn, & C. Stoel-Gammon (Eds.), Phonological development: Models, research, implications (pp. 165–189). Timonium, MD: York Press. Strauss, C., & Quinn, N. (1997). A cognitive theory of

545

Dwight Atkinson cultural meaning. Cambridge: Cambridge University Press. Street, B. (1984). Literacy in theory and practice. Cambridge: Cambridge University Press. Street, B. (Ed.). (1993). Cross-cultural approaches to literacy. Cambridge: Cambridge University Press. Stubbs, M. (1996). Text and corpus analysis. Oxford, UK: Blackwell. Tannen, D. (1986). That’s not what I meant: How conversational style makes or breaks your relationships with others. New York: Ballantine. Tannen, D. (Ed.). (1993). Gender and conversational interaction. New York: Oxford University Press. Tannen, D., & Saville-Troike, M. (Eds.). (1985). Perspectives on silence. Norwood, NJ: Ablex. Traweek, S. (1988). Beamtimes and lifetimes: The world of high energy physicists. Cambridge, MA: Harvard University Press. Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University Press. Vygostky, L. S. (1981). The genesis of higher mental fuctions. In J. V. Wertsch (Ed. & Trans.), The con-

cept of activity in Soviet psychology (pp. 144–188). Armonk, NY: M. E. Sharpe. Wenger, E. (1998). Communities of practice: Learning, meaning, and identity. Cambridge: Cambridge University Press. Wertsch, J. V. (Ed.). (1985). Culture, communication and cognition: Vygotskian perspectives. Cambridge: Cambridge University Press. Wertsch, J. V. (1998). Mind as action. Oxford: Oxford University Press. Wertsch, J. V., & Stone, C. A. (1985). The concept of internalization in Vygotsky’s account of the genesis of higher mental functions. In J. Wertsch (Ed.), Culture, communication and cognition: Vygotskian perspectives (pp. 162–179). Cambridge: Cambridge University Press. Zukow-Goldring, P., & Ferko, K. R. (1994). An ecological approach to the emergence of the lexicon: Socializing attention. In V. John-Steiner, C. P. Panofsky, & L. W. Smith (Eds.), Sociocultural approaches to language and literacy: An interactionist perspective (pp. 170–190). Cambridge: Cambridge University Press.

Modern Language Association Publication Prizes: Biennial Competitions 2003 The MLA announces deadlines for submissions for its 2003 Biennial Prizes Competitions. For more information contact the Office of Special Projects: (646) 576-5141; [email protected]. MLA Prize for a Distinguished Scholarly Edition: For single or multiple volumes. At least one volume must have been published in 2001 or 2002. 1 May 2003 Morton N. Cohen Award for a Distinguished Edition of Letters. For single or multiple volumes; at least one volume published in 2001 or 2002. 1 May 2003 Aldo and Jeanne Scaglione Prize for Studies in Slavic Languages and Literatures: For a book on the linguistics or literatures of the Slavic Languages. 1 May 2003. Aldo and Jeanne Scaglione Prize for a Translation of a Scholarly Study of Literature: For a translation into English of a book-length work of literary history, criticism, theory, or philology. 1 May 2003. Lois Roth Award for a Translation of a Work of Literature: For a translation into English of a booklength literary work. 1 April 2003. Aldo and Jeanne Scaglione Prize for Italian Studies: For a work on any phase of Italian literature or culture or comparative literature involving Italian. 1 May 2003.